In an 1893 article in The Chinese Recorder, the leading Protestant journal in China, Canadian missionary George Hartwell described Chengdu, the capital of Sichuan province, as a ‘beautiful city’ whose ‘Cosmopolitan nature’ made it an ideal centre for proselytization in western China.Footnote 1 Compared to the Catholic Church, which had operated continuously in Sichuan since the late seventeenth century, Protestant missionaries were newcomers.Footnote 2 The first Protestant mission in Chengdu opened in 1881 and, by 1895, the number had grown to four.Footnote 3 That same year, however, Hartwell was forced to hide in the house of his neighbour as a mob of anti-Christian rioters ransacked the missionary hospital nearby. During a brief lull in the riot, the city prefect and the county magistrate arrived at the scene and dispersed the crowd. Yet as soon as the officials left, ‘the mob rebegan [sic] their work of destruction’, and ‘signs of receiving official help’ vanished.Footnote 4 In two days, every Catholic and Protestant establishment in Chengdu was destroyed.Footnote 5
Anti-Christian violence soon reverberated across the province. Rioters destroyed fifty mission properties and forced most Chengdu missionaries to flee Sichuan.Footnote 6 The Qing government agreed to pay hefty indemnities for foreign losses, including almost 1 million taels of silver to the Catholic Church alone.Footnote 7 More notably, the aftermath of the riots saw the dismissal of multiple Qing officials, including several county magistrates, one middling prefect and intendant, and the governor-general of Sichuan, Liu Bingzhang.Footnote 8 The Chengdu riot marked the first time that a high-ranking territorial administrator was subjected to permanent degradation (gezhi yongbu xuyong) over anti-missionary violence, the highest level of administrative sanction below a criminal penalty.
The Chengdu riot formed part of a broader phenomenon of legal contestation over the foreign presence in China. Historians have tended to place such conflicts in the context of tensions over extraterritoriality and consular jurisdiction.Footnote 9 This article focuses instead on the indigenous protection of foreign persons and efforts by Western powers to discipline protective institutions of the Qing state. An emphasis on protection broadens the purview beyond treaty ports and their consular courts to the vast interior of China, where Western–Chinese legal politics were no less present. In the absence of full colonial control, foreign powers had to depend on local Qing authorities to ensure adequate protection for foreign individuals. As John Westlake, a nineteenth-century authority on international law, recognized, ‘the maintenance of [consular] jurisdiction must depend in the long run on the support which the local governments gave them pursuant to the conventions’.Footnote 10 The focus on protection, then, highlights the role of non-Western states as the unequal co-producers of the treaty regime.
This article examines the events and aftermath of the 1895 Chengdu riot through an analysis of missionary writings, Qing government records, and Western diplomatic communications. It finds that legal politics shaped Qing–Western interactions before, during, and after the riot.Footnote 11 Throughout this process, local authorities emerged as key actors who strategically extended and withheld protection. Away from the immediate reach of Western military power, Chengdu officials exercised considerable discretion in undermining their treaty obligations and evading accountability. They deployed indigenous legal strategies in interactions with Chengdu missionaries that blurred the boundary between indigenous and international law, as well as between legal and extralegal actions. As in colonial contexts, the subversion of law and recourse to violence were not deviations from but continuations of legal politics.Footnote 12
The Chengdu riot prompted foreign efforts to establish a mechanism of accountability for Qing failures of protection. This initiative echoed British attempts in the first half of the nineteenth century to create an imperial ‘middle power’ that could impose control on unaccountable colonial jurisdictions.Footnote 13 In the Qing empire, Western pressure focused instead on compelling central authorities to discipline territorial administrators through publicized punishments. While extraterritoriality exempted foreigners from the jurisdiction of a Qing state characterized in Western legal writings as either lawless or despotic, protection required the effective and accountable extension of the same state power to foreign individuals.Footnote 14 This article argues that the punishment of Liu and other officials for the Chengdu riot represented a pivotal moment in the protection of foreigners in China. The Qing government was forced, under the threat of British warships, to hold a high-ranking official accountable for anti-missionary violence and to attenuate its practice of shielding officials from foreign demands of punishment. Qing–Western interactions surrounding the Chengdu riot provided an important precedent for the resolution of the 1900 Boxer Rebellion, which featured punishments of officials on a major scale.
The debate about accountability and protection connects the discourses of international jurists with the practices of both Western empires and indigenous states. The latter group was simultaneously denied full membership in a Western-dominated international society and forced to assume the obligations of their unequal participation. The effective protection of foreigners became an important measure of authority and legitimacy for these states while providing Western powers with a justification for intervention. The Chengdu riot illuminates the process by which various levels of Qing bureaucracy were drawn into constructing and maintaining the conditions of China’s unequal incorporation into international society. In this article, protection emerges as a multifaceted phenomenon, both an instrument for the projection of imperial power and a malleable requirement of treaty systems that engaged different non-Western states, mediating and reinforcing the asymmetrical international relations in the nineteenth century.
Protection in non-Western states
In the late nineteenth century, non-Western states such as China, Japan, and Siam occupied an intermediate position in relation to international law, one that justified differentiated forms of protection. International jurists had come to conceptualize the family of nations in exclusionary, Eurocentric terms, according only partial recognition and rights to non-Western polities variously described as ‘differently civilized’, ‘semi-civilized’, or ‘semi-barbarous’.Footnote 15 The Ottoman empire, though formally admitted to international society in 1856, was considered as an ‘anomalous’ case that warranted continued unequal treatment.Footnote 16 While non-Western states were thought to possess some degree of state organization, their non-European institutions failed to fully meet the standard of civilization, defined in part as the state protection of life, property, and religious freedom, especially those of foreigners.Footnote 17 Extraterritoriality functioned to protect foreign individuals from indigenous systems of justice until these states could acquire full international personality through the ‘progress of civilisation’ and Westernizing reforms.Footnote 18 In different non-Western states, the aspiration to attain the civilized status and abolish extraterritoriality drove administrative and legal institutionalization based on the European model.Footnote 19
While extraterritoriality withdrew foreigners from indigenous authority, non-Western states nonetheless assumed the obligation to protect their lives and property. These states were understood to possess international personality in some aspects but not in others, as defined in their interactions with fully civilized powers.Footnote 20 Fitzmaurice has argued that the admission of non-Western polities into international law entailed duties whose fulfilment could be demanded by Western powers.Footnote 21 When urging the Qing government to punish officials who failed to protect foreigners, Charles Denby, the American minister at the time of the Chengdu riot, described China as ‘to all intents and purposes a member of the family of nations’.Footnote 22 In this instance, partial incorporation furnished not the rights of sovereignty but the duty of disciplined protection. As government effectiveness came to be identified with protecting ‘the individual rights of those who already had (emphasis in the original) such rights (citizens of existing, i.e. primarily European and American, states)’, failures of protection disproportionately exposed non-Western states to foreign intervention.Footnote 23 Latin American states with full international personality frequently faced Western interventions in the name of protection.Footnote 24 Western powers imposed a higher degree of responsibility and harsher demands for restitution on these ‘semi-civilized or backward countries’, which were made not only protectors but guarantors of foreigners’ safety.Footnote 25 In contrast, the repeated failures of the United States to address nativist violence against Chinese, Mexican, and Italian nationals did not fundamentally call the effectiveness or civilization of its government into question.Footnote 26 For non-Western states, incorporation into the international society became another mechanism to reinforce their de facto subordination.
For Western powers, the duty to protect could turn its lapses into grounds for action in the extra-European space. Benton and Clulow have identified the broader pattern in which the ‘protection over subjects by imperial powers (“inside” protection) sometimes served to thrust imperial influence and power into new territories … (“outside” protection)’.Footnote 27 Protection emergencies, defined as ‘calls to shelter imperial subjects from harm’, projected imperial power to regions of uncertain control.Footnote 28 In the aftermath of the Chengdu riot, however, it was the non-Western state that came under pressure to provide effective protection by asserting its authority over local jurisdictions. The local enforcement of protection created a range of subversive possibilities that Western powers sought to subdue through the regular and disciplined institutions of the Qing state. Protection as a responsibility of the non-Western state accentuates the importance of ‘indigenous collaborative mechanisms’ in different systems of Western domination, which provided constrained forms of agency to non-Western actors.Footnote 29 The relationship between state authority and protection was not only theorized by European political thinkers but enacted through the messy politics of the colonial and ‘semi-civilized’ world.Footnote 30 The outcome of the Chengdu riot emerged from a multisided process that involved local and central Qing officials, foreign diplomats, and Protestant missionaries, each pursuing their distinct and often conflicting agendas.
Christianity inland: Missionary presence and legal ambiguities
Since the Treaties of 1858, the Qing government had formally agreed to the legal toleration of Christianity, the protection of missionaries and their converts, and the missionary right to proselytize in the interior.Footnote 31 The expansion of the missionary body and the Chinese Christian community produced conflicts with Qing officials, the gentry, and the local population.Footnote 32 Christianity-related disputes, known as missionary cases (jiaoan), proliferated following the treaties.Footnote 33 While Protestant missionaries in Sichuan remarked on the local population’s friendliness toward their presence, the province recorded 127 major missionary cases between 1861 and 1910.Footnote 34 Prior to the growing Protestant presence, Sichuan had been a hotbed of anti-Catholic activities.Footnote 35 The late nineteenth-century growth of Protestant missions – with 1,272 missionaries arriving in China between 1887 and 1897 – heightened the likelihood of conflict.Footnote 36
Anti-Christian attacks were usually organized by the local gentry. The overlap between the Qing officialdom and the gentry class, combined with the gentry’s essential role in late Qing local governance, made local authorities reluctant to fulfil their treaty obligations; in some cases, the officials actively abetted the disturbances.Footnote 37 By the time of the Chengdu riot, missionaries had become accustomed to gentry hostility and official indifference during riots. An 1893 entry in the Chinese Recorder observed, ‘the riot is instigated by some of the literati and gentry … there is great probability that the local officials knew of the likelihood of a disturbance … [and the] local officials may or may not have aided and abetted the disturbers’.Footnote 38 The unreliability of Qing authorities meant that protection was less a straightforward implementation of treaty rights than an open-ended process of negotiation and contestation. To secure their status, missionaries took matters into their own hands through a combination of extralegal manoeuvres and diplomatic appeals.
The right of Protestant missionaries to acquire properties in interior locations such as Chengdu was a source of legal ambiguity. The 1860 Sino-French Convention allowed Catholic missionaries to rent and purchase properties in all provinces. This provision was duplicitously interpolated into the non-authoritative Chinese version of the treaty but was officially validated by the 1865 Berthemy Convention, which Beijing accepted under French pressure.Footnote 39 The British and American governments initially declined to claim this right for their citizens and opposed permanent missionary residence beyond treaty ports.Footnote 40 The objection of the British government stemmed from the difficulty of securing treaty implementation.Footnote 41 Rutherford Alcock, the British minister, admitted in 1869 that inland missionary residence was ‘entirely contingent on the disposition of the Chinese local authority to promote or oppose it’.Footnote 42 In further comments, Alcock contended that, without the cooperation of Qing authorities, protection of missionaries could only be obtained ‘by a resort to force’, but even unarmed interventions by foreign consuls would be ‘fatal to any native executive powers’ if too frequent.Footnote 43 The British position recognized the indispensable role of Qing authorities as both the interlocutor and enforcer of foreign demands.Footnote 44 Western reliance on uncertain Qing implementation of treaties consequently limited the rights theoretically available to the missionaries.
Diplomatic caution did not prevent Protestant missionaries from establishing a permanent presence in the interior. Invoking the most-favoured-nation clause of the American and British treaties, Protestant missions claimed the same right as the Catholic Church to acquire property. To circumvent local opposition, missionary properties were usually held through a Chinese Christian or rented with a ‘perpetual lease’.Footnote 45 Hartwell recalled that the Canadian mission in Chengdu obtained its premises through the latter means because ‘foreigners could not buy property outside Treaty-Ports’.Footnote 46 In this sense, neither Qing authorities nor the missionaries took the text of the treaty as static or definitive. Both sides probed the limits of missionary presence and relitigated its terms as actors attuned to the contentious issues of rights, status, and control.
These disputes could escalate into anti-Christian violence and draw the involvement of Western powers.Footnote 47 Over time, the British and American governments converged on the view that established missionary properties in the interior, acquired with customary local permission, were entitled to the protection of Qing authorities, even though the right to missionary property acquisition itself remained unrecognized.Footnote 48 In 1888, the US Secretary of State suggested that ‘it was inevitable that the restricted limits of residence and business prescribed in those treaties should be extended’.Footnote 49 Denby, the American minister, concurred: ‘If the foreigner can procure toleration in any locality … he, by degrees, may acquire vested rights, which his own government and the [Qing] Government also are bound to secure to him if attacked.’Footnote 50 Missionaries thus stood at the forefront of extending treaty limits. They secured residence in the interior not as a ready-made right but through the accumulation of formally unsanctioned acts of settlement.
The consent of local Qing authorities still formed a component of the process of missionary property acquisition. In addition to the Berthemy Convention, the Qing government mandated local authorities to punish persons selling property to Christian missions without official authorization.Footnote 51 By imposing the requirement for official permission on sellers, who were Qing subjects, local authorities regained a measure of control over foreigners otherwise exempt from Qing jurisdiction. This condition was accepted by both foreign legations and, more begrudgingly, the missions. Writing on the subject in 1889, Gilbert Reid advised his fellow missionaries to ‘first consult with the officials, and at least secure the general sanction’, though Reid admitted that this requirement had ‘frequently been neglected by missionaries’.Footnote 52 In Chengdu, Canadian missionaries chose to accommodate local demands. Virgil Hart, the American head of the mission, had to ‘surrender’ a purchased property and avoid building foreign-style houses on the advice of local officials.Footnote 53 The experience of the Canadian mission suggests that adaptations to local requirements could coexist with, and were perhaps necessitated by, the broader assertion of the right to inland residence. Through these accommodations, missionaries secured a negotiated but legitimate status in Chengdu while imposing a long-term obligation of protection on local authorities.
By 1895, the Qing’s ability to contest the terms of missionary presence had diminished. In October 1894, the Zongli Yamen, the central-level Qing institution in charge of foreign affairs, instructed provincial governors that the seller was no longer ‘bound to give notice to the local authorities of his intention to sell’, effectively removing the requirement for official permission.Footnote 54 This concession was obtained by the French minister during the First Sino-Japanese War and prompted objections from governors who stood to lose control over missionary properties.Footnote 55 In Sichuan, provincial authorities issued a proclamation that ignored the Yamen’s directive and reaffirmed the requirement for official permission.Footnote 56 Missionaries and Western diplomats would later identify this renewed attempt to regulate missionary property as a main cause of the riot.Footnote 57 The controversy over the proclamation reveals Liu’s willingness to circumvent central-level directives and the fraught position of the Yamen caught between foreign pressures and provincial resistance. The interplay between the centre and the province, compliance and evasion, and foreign demands and local enforcement anticipates patterns of Qing–Western interactions over the riot.
Official sanction of missionary residence did not solve the problem of unreliable protection by local authorities. Despite the belief of Chengdu missionaries that the danger of riot was minimal ‘in a city with so many officials’, anti-missionary incidents were rife in ‘higher administrative centres’.Footnote 58 Although extraterritoriality presented a challenge to Qing authorities following the riot, it did not produce a straightforward shift toward the kind of modernizing reforms often associated by scholars with interactions between international law and non-Western states. In the Chengdu case, Western diplomats and ministers of the Yamen relied on networks of territorial administration that justified extraterritoriality in the first place. Western efforts focused instead on ensuring that existing institutions could adequately fulfil their treaty obligations. By bringing undisciplined local authorities into compliance, reforms in the treatment of foreigners could occur without wrestling with the issue of extraterritoriality.
Unaccountable authorities: Reality and fear
Analysing the events of the Chengdu riot is problematized by the conflicting narratives of missionaries and local authorities. Each side had a vested interest in either incriminating or exonerating the behaviours of Qing officials. The contradictory motives behind these narrative constructions underscore ‘the impossibility of truly resurrecting the experienced past’, as Cohen has suggested in the context of the Boxer Rebellion.Footnote 59 Nevertheless, the triangulation of different accounts can reveal discrepancies and consistencies that offer insight into the actions, concerns, and the agenda of the actors. Foreign and Chinese accounts, for instance, concurred on the basic sequence of events. On 28 May 1895, a mob attacked the Canadian missionary compound. Over the next day, rioters destroyed the premises of other Protestant and Catholic missions, forcing all foreigners to take shelter in the county magistrate’s yamen. Disturbance faded on 30 May, when Governor-General Liu Bingzhang adopted more suppressive measures.Footnote 60
Foreign scrutiny centred on Liu, who had long been accused of harbouring anti-foreignism. His bureaucratic career reveals an administrator familiar with using legal and extralegal strategies to restrict Christian influence.Footnote 61 While serving as the administrative commissioner in the province of Jiangxi, Liu had access to legal experts who could help officials avoid accountability by manipulating reports of anti-Christian actions. After a county magistrate executed eighteen Chinese Christians, Liu warned him that the executions would provoke missionary protests. On the advice of the governor’s private legal secretary, the magistrate falsified the testimonies and reported the deaths as resulting from ‘armed fighting’ (xiedou).Footnote 62 After Liu was appointed to the governorship of Jiangxi, he was informed that his predecessors had assigned officers to covertly monitor missionaries and escort them to the yamen when riots inevitably broke out. Missionaries would then be advised to leave the province.Footnote 63 This tactic aimed less at preventing anti-missionary violence than at protecting missionaries from immediate physical harm while leveraging such violence to expel them.
In Sichuan, the manipulation of legal charges was central to Liu’s effort to undermine the Catholic Church. Liu started his tenure as the governor-general in 1886 by securing the execution of a Chinese Catholic magnate whose guards killed eleven rioters during an anti-missionary incident.Footnote 64 Although the magnate’s action was arguably in self-defence, Liu ordered his legal secretaries to frame the killings as a case of armed fighting. The charge warranted the beheading of the culprit. Despite objections from the French minister and other high-ranking officials, Liu reported the execution through the recently built telegraph system before the French minister could intervene.Footnote 65 For the Chengdu riot, Liu could draw on this repertoire of strategies – the selective provision of protection, legal manipulations, and the strategic use of telegraphic communications. They were employed to undermine the protection of missionaries and to shield officials from the scrutiny of both the central government and Western powers. In turn, the perception of local unaccountability would fuel foreign anxieties that Qing officials had found a way to subvert the treaty regime.
During the riot, the protection provided by local officials was ineffective and unreliable. Missionaries argued that the riot could have been suppressed with adequate force. Henry Cady of the American Methodist mission believed that a force of twenty armed men would have controlled the mob attacking his mission.Footnote 66 Instead, on both days of the riot, foreigners were largely left to fend for themselves before being taken to the yamen on the night of 29 May.Footnote 67 After the first day, the remaining Protestant missions received additional runners from the yamen, who also failed to provide effective protection. Cady testified that they joined the looting after the mission compound was breached, while Joseph Vale of the China Inland Mission noted that runners did not use their weapons on rioters.Footnote 68 Although Chengdu was the seat of multiple yamens capable of extending protection, foreign and Chinese accounts agreed that the magistrate of the Huayang county, whose jurisdiction included the city proper, was the primary official to respond to the emergency.Footnote 69 An American inquiry later concluded he was the only official to ‘make any energetic effort to suppress the riot’.Footnote 70 The magistrate claimed that his superiors also ‘personally come to the scene to supervise’ protective efforts on 28 May.Footnote 71 The reference to the early involvement of superior officials could be an act of provincial self-preservation in the face of central and foreign scrutiny. Qing governors wielded extensive power over their subordinates, and Liu’s unusually long tenure in Sichuan, would have ensured that the magistrate’s narrative aligned with provincial interests.Footnote 72
The lack of effective protection during the riot resulted from Liu’s deliberate decision. In his report to Beijing, Liu attributed the failure of protection to extraordinary circumstances beyond official control, stressing that the simultaneity and intensity of anti-Christian attacks on 29 May overwhelmed local authorities.Footnote 73 However, Liu’s own report indicated he only imposed martial law after every missionary property had been destroyed and anti-Christian placards threatened further attacks on Chinese Christians.Footnote 74 Qing officials would later confirm that the governor-general withheld troops. The magistrate testified that Liu personally refused his request for troops in the early morning of 29 May.Footnote 75 The refusal was corroborated by Liu’s sympathetic successor as the governor-general and a more hostile supervisory official in Beijing.Footnote 76 From the outset, the magistrate lacked adequate resources to provide protection, even as his hobbled attempt to suppress the riot was presented as evidence of a good-faith official response. Compared to what he learned in Jiangxi, Liu’s inaction in the first two days of the riot demonstrated a greater willingness to risk harm to the missionaries. Nevertheless, the broader patterns behind the selective provision of protection remained consistent. By withholding protection and then extending it when escalation appeared likely, the governor-general could manage the scale of violence, enable the removal of the missionary presence, and present the actions of Chengdu authorities as consistent with treaty obligations. This strategy allowed the riot to run its course while preserving plausible deniability for officials.
Chengdu officials were also responsible for the legitimation of anti-Christian rumours. Prior to the riot, missionaries had been accused of kidnapping and killing children. On the morning of 28 May, a placard appeared in the city warning that foreigners were extracting children’s bodily ‘oil’.Footnote 77 In missionary cases, such rumours frequently preceded violence, acting as a tool for the gentry to mobilize popular anti-Christian actions.Footnote 78 The specific accusation of child kidnapping appeared in more than 20 per cent of major rumour-based missionary cases.Footnote 79 It was closely associated with the crime of ‘dismembering a person to extract vitality’ (caisheng zhege), which as a ‘violation of fundamental taboos’ was one of the gravest offences in the Qing Code.Footnote 80 Yang has argued that the gentry used this association to ‘rationalize anti-Christian actions on a legal basis’.Footnote 81 In Chengdu, officials directly validated the authenticity of the child-kidnapping rumour. On 29 May, Zhou Zhenqiong, the intendant overseeing law enforcement in the city, posted a proclamation declaring that the authorities ‘have obtained clear proof that foreigners deceive and kidnap small children … [and the officials] certainly will not be lenient with them’.Footnote 82 This official confirmation of anti-Christian allegations, publicized while the riot was underway, further inflamed popular sentiments. Even Liu tried to halt the proclamation and denied to Beijing that it contained anti-missionary messages.Footnote 83 Prior to the 1870 Tianjin massacre, the appearance of similar allegations in official proclamations exacerbated anti-Christian tensions.Footnote 84 Zhou’s notice would have served a comparable purpose in Chengdu, mobilizing the rioters by constructing a legitimate veneer for their cause.
Liu also seized on the accusation of vitality extraction, framing it as the official explanation for the Chengdu riot. On 29 May, Liu twice telegraphed the Zongli Yamen that the riot was provoked by missionaries who dragged two people into the Canadian hospital. According to Liu, an enraged mob then broke into the premises and unearthed a tin box containing a boy who had been kidnapped into the compound, force-fed a ‘black-powdered potion’, and rendered speechless.Footnote 85 Liu’s reports transformed popular rumours into bureaucratic narratives that appeared before the Yamen as credible information. An incredulous Yamen nevertheless questioned Liu’s account: ‘If the church’s tin box did contain the boy, he would have suffocated before the box was found.’Footnote 86 The reply suggests that anti-Christian rumours were not universally believed within the Qing officialdom. Zhang Zhidong, the governor-general of Huguang, strongly reproved the vitality extraction rumour in 1898, indicating similar scepticism among provincial governors.Footnote 87 Despite the countervailing attitudes of his bureaucratic peers, Liu doubled down on the authenticity of the accusation. On 1 June, provincial authorities stressed to the Yamen that ‘facts about [missionaries] concealing and harming children are true and cannot be manipulated’.Footnote 88 The provincial insistence on the accusation points to the central role it would subsequently play in Chengdu officials’ management of the riot’s aftermath.
To both Beijing and the foreigners, Chengdu officials used allegations of missionary wrongdoing to claim lesser responsibility for their failure to provide protection. In a secret report, Liu contended that foreigners had long ‘bullied’ the local population, whose grievances were triggered into a riot by the discovery of the child.Footnote 89 By framing vitality extraction as the catalyst of an uncontrollable mob, Liu deflected the focus on the incident from the failures of the authorities to purported foreign misconduct. This position was echoed by other officials. The city prefect told members of the Canadian mission, ‘The discovery of the child was witnessed by countless people, which made it impossible for high officials to intervene on your church’s behalf.’Footnote 90 For Chengdu officials, missionary wrongdoing limited the scope and effectiveness of their response, thereby lowering the protection that should be expected from the authorities.
These assertions of missionary crimes were bolstered with staged evidence. Omar Kilborn of the Canadian mission recounted that ‘[human] heads, hands, and feet’ were paraded on the street and brought to the yamen.Footnote 91 The manipulation of human remains served a clear judicial purpose. Macauley has argued that the dead body ‘was an effective tool of legal empowerment’ used in Qing-era false accusations by disadvantaged litigants to rebalance power dynamics in their favour.Footnote 92 In Chengdu, this subaltern legal strategy was repurposed to lodge accusations against missionaries protected by extraterritoriality. Indeed, even more potent than dead bodies, the authorities possessed a living person alleged to be the victim of vitality extraction.
The child victim provided the officials with an opportunity to subject missionaries to a quasi-judicial process. While confined to the yamen, three missionaries were questioned by the prefect in front of the child. The nature of this encounter was ambiguous. The prefect used the term ‘meeting’ (huiwu) to describe the interaction.Footnote 93 Canadians David Stevenson and Omar Kilborn instead wrote of ‘being tried’ and cross-examined by the prefect, although Joseph Vale believed it was not a trial.Footnote 94 Liu likewise framed the inquiry as part of a judicial process. According to the governor-general, the child was brought out for ‘cross-examination’ (duizhi) with missionaries whose denials made it difficult to ‘reach a verdict’ (dingyan).Footnote 95 This ambiguity allowed officials to interrogate foreigners attuned to their extraterritorial privileges. In late Qing Sichuan, officials condoned false accusations in lawsuits, which allowed them to intervene in local communities.Footnote 96 In this case, the child’s in-person accusations compelled the missionaries to engage with and refute the claims in a quasi-legal process presided over by the officials, thereby bringing them under Qing scrutiny. Such accusations, along with rumours and dead bodies, typically functioned as the weapon of the weak.Footnote 97 Their adoption by Chengdu authorities, the usual powerholders in local society, reflects the juridical asymmetry created by extraterritoriality. Yet the boundary between indigenous and foreign jurisdictions proved porous. The redeployment of Qing legal charges and strategies in interactions with foreigners undermined both treaty-mandated protection and extraterritoriality. The role of local authorities as the ultimate enforcer of the treaties produced a middle space in which Qing legal practices intersected with and, at times, subverted the obligations and standards imposed by Eurocentric international law.
The ‘trial’ of Chengdu missionaries also reveals protection as a precondition for extraterritoriality in the interior. Stevenson noted that the inquiry ended when Kilborn ‘whipped out his British passport and showed the strong lines in which it said if we committed any wrong we were to … be sent with due care to the nearest British consul’.Footnote 98 This confident assertion of foreign inviolability belied the missionaries’ pleas for official intervention during the riot. It also understated the full extent of the proceedings, during which the Canadians were made to answer additional allegations of misconduct.Footnote 99 At the time of this encounter, the foreigners were safely sheltered within the yamen. Only after the authorities extended their protective power to foreign persons were the missionaries able to challenge the officials’ effort to circumvent extraterritoriality. In this sense, the physical protection extended by Chengdu authorities preceded the extraterritorial protection of foreigners from the same indigenous power. Indeed, even maintaining consular jurisdiction in treaty ports relied on Qing cooperation. In the eastern Sichuan city of Chongqing, the nearest treaty port, successive British consuls had looked to city authorities for protection.Footnote 100 During the 1895 riots, the consul attributed the absence of trouble ‘solely to the energetic action’ of a friendly intendant and urged the British minister to press for the official’s retainment when he was called to Beijing.Footnote 101 In both cities, the intention and initiative of the local officials determined the effectiveness of the treaties.
The uncertainty of treaty compliance was further complicated by Liu’s control of telegraphic communications. By the time of the riot, telegraph lines allowed the rapid transmission of information from Chengdu, but their control remained in Chinese hands.Footnote 102 On 29 May, Chengdu missionaries managed to send one telegram about the riot to the British consul in Chongqing, but their second telegram in the next morning was refused.Footnote 103 An element of chance determined the consul’s receipt of information. He was told that officials attempted to intercept the first telegram and that ‘had it not been for the friendliness of one of the operators no prompt news about the riot and its consequences would have reached [him]’.Footnote 104 While Chengdu officials withheld communications channels from the foreigners, they appropriated the same network to spread anti-Christian rumours. Three days after the riot, Nicholas O’Conor, the British minister, warned the Zongli Yamen that an official telegram from Chengdu propagated the child-kidnapping accusation Liu had reported to Beijing.Footnote 105 Without the chance 29 May telegram, the official version of the riot and its cause would have been the only information available to both the central government and foreign legations before early June. With the control of communications, Chengdu officials could have edged out countervailing foreign claims and bolstered their attempt to evade accountability.
The Yamen’s response to foreign concerns reinforced perceptions of an unaccountable local power. With few independent means of verification, O’Conor pressed Beijing to investigate official malfeasance. The Yamen, however, sent its directives to the very Sichuan institutions that had been the subject of foreign complaints. Predictably, Liu reported that communications problems were the result of malfunctioning telegraph lines and that no anti-Christian telegram was sent from Chengdu.Footnote 106 His replies were disproved by other Qing officials. Sheng Xuanhuai, the influential customs intendant of Tianjin, referenced a ‘secret order’ by Liu to stop foreign telegrams.Footnote 107 Further evidence came from O’Conor, who reported that the British consul in Hankou received from Chinese authorities a copy of the Chengdu telegram containing anti-Christian accusations. This confirmation prompted the exasperated British minister to call for a ‘searching inquiry’ and the public punishment of culpable officials.Footnote 108 Despite Liu’s attempts to control the flow of information, the news trickling out from the interior revealed patterns of official complicity that compromised the provincial and central government in the eyes of Western actors.
In China’s foreign community, narratives of official malfeasance fuelled fears that Qing authorities had found a way to sabotage their obligation of protection. Foreign reports of the Chengdu riot framed Liu as its direct instigator. Alicia Little, a British resident of Chongqing, wrote that Liu instructed the mob: ‘You can pull down what you like and rob what you like, but do not burn anything lest you should set fire to the neighbours’ houses.’Footnote 109 Similarly, the widely read North China Daily News commented that the conduct of Chengdu authorities towards missionaries was ‘Drive out, but do not kill’, while also reporting the rumour that Liu had declared ‘all losses would be met’.Footnote 110 In these accounts, Chengdu authorities channelled violence towards missionary properties but regulated the extent of destruction, fully aware that some form of compensation would eventually be offered. Local officials were portrayed as exploiting gradations of Western responses to protection emergencies of varying severity. They appeared to understand that while harm to foreigners might provoke significant interventions, the destruction of property could confine penalties to monetary compensation.
Missionaries began to believe in the existence of a broader Qing strategy to expel foreigners while minimizing Western interventions. F. R. Graves, the American Anglican Bishop of Shanghai, wrote of how riots generally unfolded: officials would harass Chinese sellers of property, and the gentry would organize the mob through allegations of child kidnapping. ‘But the master touch of all’, Graves concluded, ‘is the caution to the crowd to stop short of violence, not in mercy to the innocent, but to save the mandarin from getting into trouble (emphasis in the original) … and the destruction of property will accomplish the end.’Footnote 111 Missionaries feared that monetary compensation enabled unaccountable local officials to abet riots through calculated, limited acts of violence with controlled consequences. In these narratives, the existing treaty regime, rather than restraining Qing authorities, appeared to have facilitated anti-missionary riots by creating standards that could be manipulated and averted. Consequently, finding a way to rein in the arbitrary power of local officials became the priority of Western powers.
Disciplining protection: Diplomacy, gunboats, and accountability
The Chengdu riot prompted Western efforts to hold Qing officials accountable for anti-missionary violence through punitive sanctions. Benton and Clulow have argued that ‘[within] a particular encounter, meanings of protection could swing between weightily substantial and airily vague’.Footnote 112 Foreign demands to punish Chengdu officials challenged the Qing practice of shielding officials from substantial personal responsibility for failures of protection. In previous missionary cases, the Qing central government had sought to deflect similar demands through monetary compensations and harsh punishments for non-elite rioters.Footnote 113 When lower-ranking officials faced punishment, the eventual penalties were usually limited.Footnote 114 O’Conor noted that Beijing had ‘under pressure – sometimes only under great pressure … dismissed occasionally a petty mandarin’.Footnote 115 The absence of adequate punishment following the 1891 Yangzi riots, during which two foreigners were killed in central China, exacerbated foreign fears that local authorities were permitted by Beijing to challenge the treaty regime.Footnote 116 In contrast, the permanent degradation of Liu Bingzhang in an imperial decree on 29 September 1895 appeared extraordinary. As Denby, the American minister, informed Washington, ‘there was no precedent for this action’.Footnote 117 The harsh punishments of Chengdu officials marked the beginning of the turn towards increasingly stringent punitive measures from Beijing targeting higher-ranking officials in response to anti-missionary violence.Footnote 118
This outcome was shaped by protracted negotiation and contestation between Beijing and Western powers. Initially, Qing authorities adhered to the pre-existing formula of resolving missionary cases. On 26 June, Beijing ordered Liu to negotiate a monetary settlement with the Catholic mission and to ‘rigorously’ punish rioters to ‘avoid giving grounds for [foreign] criticisms’.Footnote 119 Robert Hart, an influential British administrator employed by the Qing government, believed that Beijing preferred paying the exorbitant indemnity charged by the Catholic mission to degrading Liu.Footnote 120 However, the American and British ministers rejected the monetary approach. O’Conor notified the Yamen that ‘Without such punishment of the Viceroy, if guilty, I can consider no settlement of the case satisfactory.’Footnote 121 Denby conveyed the same opinion: ‘officials who are criminal or negligent in their conduct toward foreigners must be punished, and that simple payment of damages is not enough’.Footnote 122 These statements tied the resolution of the riot to the attribution of official responsibility. The indemnity was positioned as a derivative of, rather than a substitute for, the punishment of culpable officials. Denby, in particular, operated under considerable pressure from American missionaries.Footnote 123 Both before and after the riot, the Protestant missionary community rejected the mere Qing offer of indemnities, condemning it as ‘blood money’.Footnote 124 The American minister withdrew from a British-led commission of inquiry in July following missionary criticisms of his decision as an inadequate American response.Footnote 125
Obtaining such concessions from Qing authorities proved challenging. Denby’s proposal for joint Western gunboat action to bombard Chinese towns was rejected by a State Department that preferred limited commitments in China.Footnote 126 When Virgil Hart visited him following the riot, the American minister was pessimistic about ‘securing much beyond a money indemnity’.Footnote 127 The British minister’s attempt to organize the Chengdu inquiry illustrates the difficulty of translating diplomatic demands into concrete results. O’Conor initially intended the British consul in Chongqing to travel to Chengdu and join the inquiry. However, the consul was confined to his post by the threat of riots in the city, and the Yamen ordered the intendant in Chongqing to negotiate with foreigners in the treaty port.Footnote 128 O’Conor agreed to the change of location on the condition that the provincial judicial and administrative commissioners could join the inquiry. By 17 July, he had settled for the judicial commissioner alone travelling to Chongqing.Footnote 129 More than a month after the riot, the British legation had not begun the formal process of determining official responsibility. Instead, the minister repeatedly demonstrated his willingness to accommodate Qing requests. Western powers recognized the danger inherent to this delay. The State Department noted that ‘long negotiations’ negated the ‘material and moral effect’ of punishment.Footnote 130 These difficulties meant that the eventual punishment of Liu was far from assured. Drawing a direct line from Western pressure to Liu’s permanent degradation overlooks the uncertainty and contingency of the process.Footnote 131 In the Chengdu case, a crucial contingency was occasioned by the killing of eleven British missionaries in Gutian on 1 August 1895.
The massacre in Gutian, a county in the coastal province of Fujian, transformed the punishment of Chengdu officials into an urgent diplomatic priority. In terms of foreign deaths, the massacre represented the worst outbreak of anti-missionary violence between the 1870 Tianjin massacre and the 1900 Boxer Rebellion.Footnote 132 Satisfied by proofs of Liu’s culpability, O’Conor informed London: ‘the brutal and horrible massacre of British subjects … made it clear that it would not be well to wait for the result of the [Chengdu] enquiry or to delay any longer in demanding the severe punishment of the ex-Viceroy’.Footnote 133 Governor-General Liu became, inadvertently, the subject of foreign fury for an initially unrelated missionary case. Unlike in Chengdu, the perpetrators of the Gutian massacre were members of a local religious group more hostile to the Qing state than to the missionary presence.Footnote 134 The British consul in Fujian, while holding the provincial government responsible for the failure to protect, implied that the recently appointed governor-general of the province was difficult to blame.Footnote 135 Chengdu officials, already found to be culpable in missionary narratives, emerged as more convenient and immediate targets of reprisal.
Public pressure from the missionaries and the foreign press played a considerable role in hardening the diplomatic resolve to punish Liu. Influential members of the expatriate community framed the Gutian massacre as the consequence of the diplomatic failure to punish Chengdu officials. The China Association, an influential interest group of British merchants, organized a well-attended ‘indignation meeting’ of foreigners in Shanghai.Footnote 136 John R. Hykes, an American missionary who had previously spoken on Chengdu, declared the Gutian massacre to be ‘a terrible comment upon the masterly inactivity with which the Sezchuan riots have been treated’.Footnote 137 The British chairman of the meeting similarly warned that ‘the Chengtu commission … is treated with contemptuous derision by Chinese’, and that without the punishment of culprits ‘whatever their rank may be … the lives of English people in China will be more and more endangered’.Footnote 138 These statements conveyed the growing expatriate frustration not only with Qing authorities but also with foreign diplomatic representation. Following the massacre, foreign expatriates sidestepped their legations and appealed directly to home governments and metropolitan opinion. The Shanghai meeting resolved to ‘get the facts of the position in China to the knowledge of the great American and British public’.Footnote 139 An article in the Chinese Recorder later marvelled at the unprecedented ‘sympathy and good-will’ of London newspapers toward missionary work.Footnote 140 The resulting pressure reached both the Qing and foreign officials. Li Hongzhang, a pre-eminent Qing statesman, warned the Yamen that ‘the entire [British] nation is quite indignant’.Footnote 141 Both Denby and O’Conor commented on intense expatriate criticisms of their previous actions, with the latter reporting that the ‘Missionaries have frightened my American Colleague out of his senses.’Footnote 142 By galvanizing expatriate and metropolitan attention, the Gutian massacre reinvigorated the push to enforce effective protection. Central to these efforts was the demand for Liu’s exemplary punishment, which became a focal point in subsequent negotiations.
The Qing government strenuously resisted the foreign demand as an encroachment on its sovereign authority. In August, Denby and O’Conor called for Liu’s permanent degradation from office, banishment to the frontiers, and public notice of the reasons for his punishment.Footnote 143 While the Yamen was open to punishing Liu, it rejected the proposal of banishment and suggested that the decision of degradation was an internal issue. O’Conor was told by Weng Tonghe, a prominent minister of the Yamen, that ‘the matter was one to be dealt with by the Chinese Government strictly in accordance with Chinese modes of procedure’.Footnote 144 Over the following weeks, Qing leaders reiterated their insistence on punishing Liu according to Qing disciplinary protocols, with decisions concerning banishment reserved for the throne and the Board of Punishment.Footnote 145 The terms of Liu’s penalties not only set up a clash between Qing law and Western demand, but one centring on the crucial issues of political authority and control over official careers. Weng feared that acceding to Liu’s permanent degradation could set a dangerous precedent for foreign ‘intervention in [Qing] internal governance’.Footnote 146 The Yamen instructed its minister in London to protest O’Conor’s request as ‘in contradiction to international law’.Footnote 147 The French minister likewise opposed British demands and supported China’s right to prescribe punishments on its own terms.Footnote 148 The Yamen’s resistance thus represented an effort to claim and safeguard the sovereign right available even to China as a non-Western state, articulated in terms recognizable to Westen powers through the language of international law.
The events leading to Liu’s degradation demonstrate the challenges of overcoming Qing opposition even with armed intervention. The British, French, and American ministers each claimed credit for Liu’s downfall.Footnote 149 Denby attributed Liu’s punishment to the American insistence on sending a commission of inquiry to Chengdu, although the commissioners would not leave for their destination until October.Footnote 150 It was O’Conor’s ultimatum backed by British warships that compelled Beijing to concede. Weng’s account reveals that as late as 27 September – two days before O’Conor’s ultimatum expired – the Yamen remained divided on whether to comply with the British demand.Footnote 151 The imperial decree that cashiered Liu imposed the penalty of permanent degradation but stopped short of banishment.Footnote 152 O’Conor had agreed to drop the latter on 26 September ‘to avert an appeal to force’.Footnote 153 The British minister’s willingness to soften his demand likely reflected the difficulties of projecting Western forces into the deep interior. Since the waterway into Sichuan was inaccessible to gunboats, the city of Wuchang in central China was selected as one site for naval demonstration.Footnote 154 Lord Salisbury, the British prime minister, feared that it would be ‘very barbarous’ to bombard the city for ‘the exile of a mandarin who lives somewhere else’, while naval officers warned that modern fortifications along the Yangzi posed a real threat to British ships.Footnote 155 As Robert Hart noted on 22 September, the naval demonstration ‘may end in a fiasco’.Footnote 156 Even under the shadow of force, the outcome of Western pressure remained uncertain until the final moments. The challenges of implementing gunboat diplomacy underscored the significance Western powers attached to the punishment of Liu, which was meant to introduce a regular measure of official accountability that could preclude future need for direct intervention in protection emergencies.
The terms of Liu’s punishment – permanent degradation and the publicizing of his culpability – exemplified aspects of the accountability mechanism that Western powers sought to establish. This mechanism was intended to prevent future riots by linking protection failures directly to punitive sanctions on Qing officials. As O’Conor warned the Yamen, ‘there will be no security for the lives and property of foreigners in the interior until the officials throughout the empire have been taught that remissness in protecting them will bring down punishments on their own heads’.Footnote 157 The American government similarly emphasized the necessity of attributing concrete responsibility to local authorities.Footnote 158 As Liu’s case was meant to demonstrate, sanctions would be applied according to the logic of hierarchical responsibility. O’Conor believed that the ‘severe and exemplary punishment of a high official’ could stop further riots.Footnote 159 Washington adopted a more systematic stance, demanding that responsibility for failures of protection should directly extend to ‘the viceroy or governor of the province … although his only fault may be ignorance’.Footnote 160 Accordingly, the provincial leadership could face punitive sanctions for anti-missionary violence in any subordinate jurisdiction. Liu’s punishment thus served as both a precedent and a warning. By expanding the accountability mechanism across the bureaucratic hierarchy, Western powers sought to instil a deterrent effect that would restructure the Qing approach to protection and compel local authorities to actively prevent or effectively respond to anti-missionary violence.
Diplomatic demands combined punitive sanctions with the requirement for Qing authorities to publicize the reasons for punishment. Deterrence would only work if culpable officials and their peers recognized their punishment as a direct consequence of protection failures. O’Conor believed that when the Qing government previously removed officials without disclosing ‘the real grounds for the dismissals’, it undermined the ‘moral effect’ of punishment.Footnote 161 In Liu’s case, the imperial decree’s public acknowledgement of his failure to protect missionaries was described by Robert Hart as the ‘hardest of all to swallow’ for the Qing officialdom.Footnote 162 The requirement for public punishment extended to lower-ranking officials. O’Conor pressed the Yamen for the dismissal of Zhou Zhenqiong, the intendant who validated anti-Christian rumours, and for the entry of his punishment into the provincial gazette.Footnote 163 As an important medium for the circulation of official information, the gazette was widely read by officials across a province.Footnote 164 The publication of Zhou’s culpability in such a visible channel would amplify the deterrent effect of his punishment. O’Conor ultimately agreed to drop this demand after obtaining the Yamen’s assurance that Zhou would be permanently disbarred from holding office.Footnote 165 Notably, this penalty exceeded Zhou’s formal punishment of simple degradation.Footnote 166 These negotiations over the terms of the intendant’s punishment indicate that the requirement of publicization was treated as a major component of official accountability. When the requirement was traded away, its omission warranted an increase in the severity of punishment.
The terms imposed on Chengdu officials were integrated into broader diplomatic proposals for accountability and punishment. Western powers placed the burden of enforcing these measures on the Qing central government. As a veteran American missionary observed, ‘the police of the West, whether in the form of a “gun-boat” or a plenary commission’ would not be necessary when ‘China, in all its parts … is to be reached from Peking’.Footnote 167 To the Qing central government, Denby argued that ‘the best means to prevent the recurrence of antiforeign riots’ were the imperial declaration of the missionary right to residence and property acquisition, the extension of hierarchical accountability to both negligent and malicious officials, and an increase in the severity of penalty from permanent degradation to ‘death, imprisonment, confiscation of property, [and] banishment’.Footnote 168 These terms, based on the recommendations of the Chengdu commission, represented a departure from the previous American policy of non-recognition for such a right. They reflected the diplomatic cognizance that the promotion of missionary rights would only be effective when tied to the enforcement mechanism of official accountability. The American minister noted that the punishment of high provincial officials represented ‘the most important suggestion’ made to the Yamen.Footnote 169 The British government similarly shifted its stance. Even before Liu was degraded, O’Conor had proposed, in addition to publicly punishing officials, more stringent financial indemnities and the suspension of the civil examination for official candidates in jurisdictions where riots occurred.Footnote 170 In 1898, his successor as the British minister expressed support for the missionary right to purchase property in the interior, drawing, in part, from an 1895 imperial order issued in response to the Chengdu and Gutian incidents, which had urged the protection of missions and threatened punishments on local officials.Footnote 171
The post-1895 Qing approach to anti-missionary violence did incorporate more punitive measures for officials. In May 1896, the Yamen proposed that territorial administrators – from magistrates to prefects, intendants, provincial commissioners, and governors – personally assume financial liability for riot indemnities.Footnote 172 This measure heightened the personal consequences for protection failures and held the entire chain of territorial administration, rather than just its top leadership, accountable for lapses at its lowest rung. It was adopted by the Qing government despite dissenting views within the bureaucracy, although the eventual implementation was lacklustre.Footnote 173 The principle of hierarchical accountability extended to the application of punitive sanctions. After another series of riots in 1898, an imperial decree warned provincial governors that they could be punished along with the local officials they supervised.Footnote 174 By threatening enhanced financial liability and punishments for a wider range of officials, the two measures found parallels in the demands of Denby and O’Conor after the Chengdu riot.
Nevertheless, on the crucial issue of punitive severity, the Qing government continued its practice of shielding officials. According to an official complaint in October 1895, the degradation of Sichuan officials contravened the principle of ‘non-severity’ (buying zhong), which meant that, in the past, the maximal penalty ordinarily imposed for the destruction of missionary premises was ‘demotion while retaining the current post’ (jiangji liuren).Footnote 175 The aftermath of the Chengdu riot thus underscored the need for clearer standards of official accountability. Alongside its proposal for indemnities, the Yamen submitted a separate memorial on penalties for failures of protection. The memorial distinguished between officials who ‘deliberately connived’ in riots and those who faced sudden emergencies. Those in the first category would be dealt with by the Yamen on an ad hoc basis. In the second category, officials who had provided effective protection during riots would be punished by ‘demotion in one rank while retaining the current post’. Those who failed to adequately protect foreigners would face a demotion in two ranks.Footnote 176
By regularizing the pre-existing disciplinary measures, this memorial sought to reconcile foreign demands for punishment with Qing preference for leniency. Officials in the second category were classified as having committed a ‘public offence’ (gongzui), which, under Qing disciplinary rules, referred to acts of official negligence rather than malicious intent.Footnote 177 As a result, the prescribed sanctions were significantly less severe than those demanded by Western powers. The punishment of ‘demotion while retaining the current post’ was among the less severe administrative sanctions, falling far short of the permanent degradation imposed on Liu and the criminal punishments advocated by Denby.Footnote 178 As Dong has suggested, the memorial’s measures constituted a ‘“protective” barrier’ for officials.Footnote 179 In this sense, foreign pressure achieved only partial success in reshaping the Qing approach to official accountability. Following the Chengdu riot, the Qing practice of shielding officials persisted in an attenuated form, operating through, rather than in opposition to, central discipline.
The tension between punishment and leniency was active during the 1897 Juye incident. The killing of two German missionaries provided the pretext for the German seizure of territorial concessions on the coast, jump-starting the Scramble for Concessions among Western powers.Footnote 180 In response to German demands for the severe punishment of local officials, Beijing imposed punitive sanctions in line with its 1896 policy, while enhancing the penalties to reflect the gravity of the case.Footnote 181 Nevertheless, the Yamen ‘resolutely refused’ to impose permanent degradation on the governor, only agreeing to disbar him from high-level appointments.Footnote 182 The British and American governments, despite their emerging concerns about the partition of China, approved of the governor’s punishment.Footnote 183 The Juye case illustrates the uneasy balance between central discipline and foreign pressure. While the 1896 measures proved insufficient to satisfy German expectations, they provided a flexible baseline that shaped the eventual resolution. Accountability remained a negotiated process, structured by the mechanisms and standards familiarized in the wake of the Chengdu riot.
Conclusion: Towards the Boxer Rebellion
During the 1900 Boxer Rebellion, Beijing openly sanctioned anti-foreign and anti-Christian actions. Foreign reliance on punitive measures from the imperial centre collapsed with fatal consequences for missionaries in the interior, whose survival once again depended on the varied willingness of local authorities to provide protection. In the south, provincial governors defied the imperial order and pledged to protect foreigners.Footnote 184 In contrast, the massacre of the entire missionary body in the northern provincial capital of Taiyuan seemed to confirm the worst foreign fears about official complicity.Footnote 185
Following Beijing’s capitulation in 1900, foreign powers systematized and enhanced the measures of accountability introduced following the Chengdu riot. The Boxer Protocol – the formal treaty between China and foreign powers – included punitive measures with clear antecedents in the demands made by Denby and O’Conor. It prescribed criminal penalties, ranging from execution to exile, for high-ranking officials and members of the imperial clan implicated in anti-foreign violence.Footnote 186 Foreign powers separately demanded the punishment of 142 local officials, including three who were eventually executed.Footnote 187 The Qing government also agreed to suspend civil examinations in jurisdictions where such violence had occurred and to uniformly impose permanent degradation on officials for future failures of protection.Footnote 188 Treaty mandate had dismantled the ‘protective barrier’ of lenient accountability. Nevertheless, even when Beijing was under foreign military occupation, Qing officials attempted to negotiate the terms of high-profile punishments and saved two imperial clansmen from execution.Footnote 189 For both Qing and foreign actors, the Chengdu riot provided the precedent that could be scaled up to meet the extraordinary demands of the Boxer Protocol.
The events and aftermath of the Chengdu riot also illustrate how the treaty regime operated on the ground. They revealed that the specific rights of missionaries emerged from their jockeying with both local authorities and Western legations. By settling in the interior, missionaries placed themselves beyond the reach of foreign gunboats and within the uncertain protection of Qing officials. In turn, the projection of Western power in response to protection emergencies assumed a paradoxical form: the enforcement of treaty obligation through the imposition of greater central discipline over unruly jurisdictions. The Qing central government became another instrument of Western power, pressed into ensuring the protection of foreign individuals privileged by the unequal relations of international society.
More broadly, the Chengdu riot reveals the conditional and coercive nature of international law for the non-Western state. To the extent that the Qing empire’s partial international personality entailed a duty to protect, the Boxer Rebellion and the spectacular failure of protection furthered the process of intensified Western control, the decline of China’s international standing, and the erosion of its sovereign rights.Footnote 190 The threat of coercion and colonial domination thus shaped the intermediate and precariously shifting position of the non-Western state in relation to international society. For the duty-laden non-Western state, the obligation to protect – or breach thereof – served as a perilous conduit through which the ‘inside’ protection of foreign individuals could be transformed into the ‘outside’ protection of imperial intervention and domination.
Acknowledgements
I am grateful to Lauren Benton and the participants of the 2023 Work-in-Progress Workshop hosted by the Yale Council on East Asian Studies for their valuable feedback on earlier drafts of this article. I am also grateful to Heidi Tworek of the Journal of Global History and the three anonymous reviewers for their thoughtful and constructive engagement, which improved this article in important respects. My thanks go as well to Managing Editor Kirsten James for her support in the editorial process. Finally, I wish to thank Weiyao Li, Jia Xie, and Hongbo Tang for their generous assistance in providing access to sources.
Financial support
None to declare.
Competing interests
The author declares none.
Jingkai Liu is an incoming JD candidate at Georgetown University Law Center. At the time of writing, he was a MA candidate at Yale University’s Council on East Asian Studies. His current research focuses on the practice and discourse of judicial autonomy in twentieth-century China.
Open access
